Manual correction of semi-automatic three-dimensional echocardiography is needed for right ventricular assessment in adults; validation with cardiac magnetic resonance

Cardiovascular Ultrasound

Research

Kambiz Shahgaldi¹  Johan Holm²  Anders Roijer²  Ellen Ostenfeld³  Marcus Carlsson⁴ 

[1] Department of Cardiology, Karolinska University Hospital, Huddinge, Stockholm, Sweden;Department of Cardiology, Skåne University Hospital, Lund University, Lund, Sweden;Department of Cardiology, Skåne University Hospital, Malmö, Sweden;Department of Clinical Physiology, Skåne University Hospital, Lund, Sweden;Department of Clinical Physiology, Skåne University Hospital, Lund, Sweden;

关键词: adult;    clinical;    three-dimensional echocardiography;    magnetic resonance;    right ventricle;    volumes;    function;   

DOI  :  10.1186/1476-7120-10-1

 received in 2011-10-24, accepted in 2012-01-06,  发布年份 2012

来源: Springer

【 摘 要 】

BackgroundThree-dimensional echocardiography (3DE) and semi-automatic right ventricular delineation has been proposed as an appropriate method for right ventricle (RV) evaluation. We aimed to examine how manual correction of semi-automatic delineation influences the accuracy of 3DE for RV volumes and function in a clinical adult setting using cardiac magnetic resonance (CMR) as the reference method. We also examined the feasibility of RV visualization with 3DE.Methods62 non-selected patients were examined with 3DE (Sonos 7500 and iE33) and with CMR (1.5T). Endocardial RV contours of 3DE-images were semi-automatically assessed and manually corrected in all patients. End-diastolic (EDV), end-systolic (ESV) volumes, stroke volume (SV) and ejection fraction (EF) were computed.Results53 patients (85%) had 3DE-images feasible for examination. Correlation coefficients and Bland Altman biases between 3DE with manual correction and CMR were r = 0.78, -22 ± 27 mL for EDV, r = 0.83, -7 ± 16 mL for ESV, r = 0.60, -12 ± 18 mL for SV and r = 0.60, -2 ± 8% for EF (p < 0.001 for all r-values). Without manual correction r-values were 0.77, 0.77, 0.70 and 0.49 for EDV, ESV, SV and EF, respectively (p < 0.001 for all r-values) and biases were larger for EDV, SV and EF (-32 ± 26 mL, -21 ± 15 mL and - 6 ± 9%, p ≤ 0.01 for all) compared to manual correction.ConclusionManual correction of the 3DE semi-automatic RV delineation decreases the bias and is needed for acceptable clinical accuracy. 3DE is highly feasible for visualizing the RV in an adult clinical setting.

【 授权许可】

CC BY   
© Ostenfeld et al; licensee BioMed Central Ltd. 2012